{
"name": "searchapi_google_patents",
"description": "Google Patents results from SearchApi, as structured JSON.",
"input_schema": {
"type": "object",
"properties": {
"q": {
"type": "string",
"description": "Search query; most engines accept native operators and advanced filters."
},
"sort": {
"type": "string",
"description": "Order of the patent results.",
"enum": [
"relevance",
"new",
"old"
]
},
"country": {
"type": "string",
"description": "Comma-separated patent office country codes, e.g. US,WO."
},
"status": {
"type": "string",
"description": "Patent status filter.",
"enum": [
"GRANT",
"APPLICATION"
]
},
"inventor": {
"type": "string",
"description": "Filter by inventor name."
},
"assignee": {
"type": "string",
"description": "Filter by assignee name."
},
"page": {
"type": "integer",
"description": "Page number of the results.",
"default": 1
},
"num": {
"type": "integer",
"description": "Number of results per page."
}
},
"required": [
"q"
],
"examples": [
{
"q": "coffee"
}
]
}
}
Paste the definition into your own agent's tools array. The schema is generated from this engine's catalog entry, and the reader's configured values ride in the schema's examples. The name is ours rather than the MCP server's — it is what your agent will call — and every engine's is prefixed searchapi_.